North America’s forests used to burn a lot more than present…

So, is climate change really making wildfires worse today?

[This is Part 1 of a two-part post… make sure to check out Part 2]

Screenshot from the Fire Information for Resource Management System – August 13, 2025.

Research into burn rates across North America before 1900 shows that before then, many parts of the continent burned more so than present. The declarative title of a recent article by Sean Parks et al. confirms this: “A fire deficit persists across diverse North American forests despite recent increases in area burned”. Similarly, recent research into Canada’s historical boreal wildfire regime suggests that contemporary burn rates in some areas of the boreal tend to fall within the historical boundaries of natural variability going back to 1800 (with some notable exceptions).

Many who follow news about climate change will find this surprising: Aren’t burn areas across North America generally increasing throughout the continent due to global warming? How are we to make sense of this apparent contradiction between high rates of historical burn rates across North America prior to the 20^th Century, when the climate was cooler than present, and the current trend of increasing wildfire burn area, as the climate warms?

What explains this see-saw, and how does climate change factor into it – if at all? To help contextualize what’s going on, I read the literature and reached out to several experts in this area. Here’s what I found out…

Yes, average burn area has increased in recent decades

First, contemporary burn rates going back over roughly the last half Century or so are indeed on an increasing trend. This chart with data from Canada’s Interagency Forest Fire Centre (below) shows that the total area burned across Canada trended up from 1983 to 2024, despite a considerable amount of year-to-year variability. The figure also hints at a remarkable countervailing trend of declining numbers of fires across Canada over this same period, suggesting that the average fire size has increased substantially.

Figure produced by author.

As a paper by Hanes et al. explains, while human-caused fires have declined in frequency and are responsible for a much smaller proportion of the total area burned, lightning-caused fires now dominate burn area in Canada. These lightning-caused fires account for around 85% of large fires (fires greater than 200 hectares), and approximately 91% of the area burned leading up to 2015 (when the study was published). This is one clue that our changing climate is potentially changing the nature of wildfires in Canada: Less fires overall; greater average size of fires, on average; more area burned, on average.

A relatively similar pattern is seen in the United States: Since the 1980s, the number of wildfires has generally trended down (if not held relatively steady), while the average burn area has increased. So, yes, the trend over the last half Century or so has been one of larger fires and increasing burn area, in both Canada and the US. This period corresponds with the ‘Great Acceleration’ in global warming seen in the Post-War Years, particularly as of the 1970s (with Earth’s climate warming more than 1°C over this period).

Figure produced by author.

Global warming is a key driver

As discussed below, changes to the climate are not the only driver shaping wildfire activity in North America. However, in recent decades, it has been a key factor driving the trends discussed above. There are a few mechanisms through which global warming can contribute to increased wildfire burn area. For one, hotter and drier conditions can result in extreme fire weather, and it can extend the overall length of the wildfire season. Further, elevated temperatures increase evaporative demand (known as the “atmospheric thirst”), leading to drier fuels that are more likely to ignite and burn with greater intensity.

As Mike Flannigan – one of Canada’s leading experts in boreal wildfire dynamics – explained to me, there are three main ingredients in the “recipe for wildfire” – Fuel, Ignition, and Weather – and changes to our climate are influencing all of these in complex ways: First, as noted above, warmer temperatures are having a substantial effect in terms of drying fuels, particularly dead vegetation that responds quickly to weather conditions (such as leaves, grasses, and needles).. Because a warmer atmosphere increases evaporative demand, it can rapidly pull moisture from this class of vegetation, leading to more readily ignitable fuels.

While it is true that a warmer atmosphere can hold more water vapour (about 7% more per degree Celsius of warming), thereby also increasing the potential for heavier rainfall events; and while it is also true that total precipitation has increased overall in North America over the last Century (notably in northern and eastern regions, while the Southern and Western portions of the continent have generally seen a decrease), wildfire researcher Sean Parks explained to me that that even if precipitation has increased in some regions due to climate change, it has not offset the effect of increased evaporative demand caused by rising temperatures during the summer seasons: “It would take a lot more precipitation to offset the effects of increased evaporative demand on fuel moisture.” In fact, prior research by Flannigan and colleagues has shown that for every degree of warming, precipitation has to increase by between 5% and 15% (depending on the fuel type) to compensate for the drying caused by warmer temperatures.

Second, the number of ignitions tells an interesting story. In Canada, these have clearly been on a downward trend, mostly because of greater public awareness and other preventative measures. And in the US, aside from relatively low numbers of fire starts in the early 80s, the total number of fires appears to also be declining since the mid-80s.

It is not clear whether the total number of lightning-caused fires is increasing, or whether those fires themselves are just becoming larger. One study projected lightning strikes to increase by 12% across the continental US for every degree Celsius of global warming, and about 50% over this Century; However, another study focusing at lightning in Canada, showed a general decline in documented lightning strikes over the last quarter Century, most pronounced in Ontario and Quebec and adjacent US states. One paper by Jain et al. regarding the exceptional 2023 wildfire season in Canda suggests that there may be improved "lightning-ignition efficiency" – wherein the existence of drier fuels makes it easier for lightning to start a remote fire, even in the case of a lower total number of lightning strikes (as was the case in 2023).

Third, the weather itself is getting warmer, with annual surface temperatures across North America having increased by about 1.5°C relative to the second half of the 20^th Century mean. The figure below shows the change in average surface temperature of our “current climate” (the average annual value over the last 15 years) from the climate of half a Century ago (the average annual value over the 15 years from 1970 to 1984). Almost every part of North America has gotten warmer over this period, by one or two degrees, on average. Warming is more pronounced further north.

Figure generated at https://climatereanalyzer.org/research_tools/monthly_maps/

And so it is that – in North America at least – modern global warming has generally contributed to more extensive and more severe wildfires, thanks in large part to drier fuels, a longer wildfire season, and declining winter snowpacks. However, humans have also influenced those wildfire ingredients in other ways, and this gets us to the question of historical wildfire burn rates…

Yes, North American forests used to burn more so than present

The experts that I spoke to suggested that many North American forested areas used to burn more so than present – ‘present’ in this instance meaning the annual average of the decade preceding 2023. Tree ring-based research (or “dendrochronology”) generally confirms this, with some exceptions. For instance, a paper by Victor Danneyrolles and colleagues found that in four key zones analyzed within Canada, three showed a relative decline in the mean annual burn rates of recent decades relative to the early 19^th Century (see Figure below). One key exception was Wood Buffalo National Park, whose burn rate over the 2014-2023 period (denoted in red dot with crosshairs) was higher than the historical range of natural variability.

Figure 4 from Danneyrolles et al., 2024: https://cdnsciencepub.com/doi/epdf/10.1139/cjfr-2024-0216

Similarly, the conceptual figure below from Parks et al. shows how even though average annual burn rates (in a prototypical U.S. forest) may be seeing increases over the last half-Century, such rates nevertheless remain at levels that are significantly below the 17^th through 19^th Century baselines.

Figure 4 from Parks et al., 2025: https://www.nature.com/articles/s41467-025-56333-8

As the conceptual figure implies, fire exclusion and suppression efforts emerging sometime in the late 1800s in the United States coincided with a drop in annual burn area. As Sean Parks explained to me, this involved a whole range of actions aside from merely actively putting out fires (or “active suppression”).

Fire exclusion was also linked to colonial expansion and settlement across the continent, in a few ways: Many Indigenous nations used fire widely (for a host of reasons – to clear land, to support hunting and agriculture, for cultural reasons, and for land management, etc.), and these practices were forcibly removed or deemed illegal during the period of colonial expansion, which substantially reduced a source of more frequent (often low-intensity fires). Similarly, many forests were cut or thinned for agricultural expansion or for the lumber industry, and this also reduced wildfire likelihood, and its ability to spread. In some cases, Flannigan noted, there was an initial increase in fire during the early colonial expansion phase, as forests were sometimes cleared and the burning of debris would sometimes cause wildfires.

Then, with the expansion of roads into remote areas, the invention of the airplane, improvements in water pumping and management systems, and better monitoring – alongside a changing philosophy in the late 19^th and early 20^th Centuries which asserted that wildfires should be eliminated, fire suppression efforts became much more extensive, and more effective. Here I should note that Ellen Whitman, a forest fire research scientist at the Canadian Forest Service, told me these types of exclusion and suppression practices were less extensive in the Canadian boreal as they were in US forests; and typically occurred later in Canada than they did in the US. Yet they were nevertheless still present.

“Civilian Conservation Corps fire fighting crew, Clark National Forest, 1937.” From: https://foresthistory.org/research-explore/us-forest-service-history/policy-and-law/fire-u-s-forest-service/u-s-forest-service-fire-suppression/

The flip-side of improvements in fire suppression is that in some areas, fuels (vegetation) were able to accumulate over time. In turn, these greater fuel loads have meant that some recent fires have been more intense than perhaps they would have been in the absence of suppression efforts. Parks et al. thus conclude that even though contemporary wildfire burn rates may fall within historical ranges, the contemporary fire severity is much greater, causing greater ecological and societal harm than the “low-to-moderate-severity” “surface fires” that characterized the pre-Industrial era. While that in itself may be more a result of changes in human forest management practices than from climate change, Parks notes that global warming has been an important compounding factor in making today’s wildfires more severe, due to the increased heat and evaporative demand (as discussed above).

While Parks et al.’s schematic shows a clear framework for understanding the contemporary fire deficit within the U.S. (and corresponding shifts in the severity of today’s fires), the boreal experts I spoke with (including co-authors of the Parks et al. study) noted that the situation North of the 49^th Parallel is diverse and complex, and that there is considerable regional differentiation. That is, the fire deficit is more representative of U.S. forests than the boreal. Whitman noted to me that “there is much less data” about whether contemporary burn rates in the boreal are at a deficit compared to the pre-Industrial era. Similarly, she noted that the 2023 fire season was so extensive (burning 4% of Canada’s forest area) that it would be “very surprising if this kind of thing happened often in the past.” That is – if it did – one would expect Canada not to have remained a largely forested country! As Flannigan added (referring to pan-Canadian burn areas in 2023, 2024 and thus far in 2025 combined): No three year period in our modern record has seen more than 5 million hectares burned in each year.

[This is the end of Part 1… please check out Part 2 when it comes out tomorrow!]